AIMI Center Year In Review 2025
Dear colleagues,
Ten years ago, most medical AI research focused on “supervised” training methods that required costly data labels produced by medical experts. Today, the focus has shifted. Many labs are working on more scalable “self-supervised” training methods like those employed for commercial language models.
Because these methods don’t require labels, we can train on massive troves of high-quality health care data to produce medical “foundation models.” These systems learn a great deal about a specific data type and can produce more accurate systems for both common and rare conditions. We are seeing truly impressive advances across a wide array of medical specialties.
As a result of this prodigious progress, there is high interest in AI across Stanford, with new AI entities and initiatives blooming everywhere. We look forward to supporting these new partners, whether in recruiting faculty, students, and staff, marshaling support for grant applications, or providing outlets to feature their work.
For example, the recent cluster hire search in the School of Medicine showed broad institutional commitment to AI. We look forward to supporting all the outstanding AI scientists that are beginning to arrive on campus as a result.
AIMI remains dedicated to supporting outstanding AI research across the university through a range of programs and shared resources. These include our internal grant programs, research cloud credit support made possible through our industry partnerships, and other forms of research enablement designed to help investigators advance promising ideas and accelerate progress. Together, these efforts reflect AIMI’s role in strengthening the research infrastructure needed to support a rapidly growing and interdisciplinary health AI community.
This fall, we held our first AIMI Academic and Industry Summit, bringing together Stanford faculty, researchers, and industry partners for a dynamic forum of two-way exchange. The summit created space to share ideas, explore emerging directions in the field, and build mutual understanding around how we can work together to responsibly advance AI in health and medicine. We plan to make it an annual event to be held consecutively with our AIMI Symposium during Stanford’s Health AI Week, June 1-5, 2026. We are also grateful for the engagement of our industry affiliates and partners, whose real-world perspectives contribute to the richness of the AIMI community.
This year also marked significant growth across our education programs, with record participation across multiple offerings. Our annual AIMI Symposium, along with the AI+HEALTH conference co-hosted with the Stanford Institute for Human-Centered AI (HAI) and Stanford CME, reached new audiences, and we launched a new AIMI Pediatric AI Symposium focused on the unique opportunities and challenges of AI in pediatric care. We also saw unprecedented interest in AIMI’s high school programs, with thousands of students worldwide seeking to engage with AI in health and medicine. Through our ongoing summer research internship and bootcamp, we trained 100 students, and this year we launched a new academic-year internship to provide deeper, longer-term research experiences that extend AIMI’s commitment to building early pathways into health AI.
Because of the vision of former Radiology chair Sam Gambhir, AIMI resides in the Department of Radiology while it serves investigators from over 20 different departments across Stanford. We are delighted to welcome a new chair of Radiology, Professor Umar Mahmood from Harvard/MGH, who has identified AI as one of his priorities.
In response to the changing landscape of AI, we initiated a strategic planning process this past year, to solidify our goals for the coming years. We compared our strengths and weaknesses to related programs at Stanford and across the country. We then asked key Stanford luminaries and stakeholders to help us refine our mission and identify new opportunities. Finally, we conducted workshops to identify our top strategic priorities for the AIMI Center in the years to come. These key priorities are:1. Data Enablement: Facilitate responsible, scalable, and equitable access to high-quality, multi-modal health data to advance AI research and clinical innovation.2. Education and Engagement: Expand AI education to equip clinicians, researchers, students, and executives with the skills needed to integrate AI into healthcare practice.3. Research Methods for AI Translation: Advancing cutting-edge methods to develop, validate, refine, and deploy AI models into clinical practice.
Our thanks go out to everyone who devoted time and expertise to the process. This initiative served as a unique opportunity to build an adaptive organization that supports the strengths of the growing health AI activities across campus. In the coming months, we will be aligning our leadership structure with these priorities.
As we welcome the new year, all of us at the AIMI Center extend our sincere appreciation for the scholarly work you do, and for your continued partnership. We wish you all a happy and healthy 2026.
Sincerely,
Curt Langlotz
Director, Stanford AIMI